An Easy way to deploy app (alternative of GKE)
Cloud Run brings "serverless" development to containers and can be run either on your own Google Kubernetes Engine (GKE) clusters or on a fully managed PaaS solution provided by Cloud Run. You will be running the latter scenario in this lab.
GAE see https://github.com/QueenieCplusplus/GAE
(1) create a Docker Image for App
(2) move the Image to Cloud Storage Bucket for Cloud Build.
(3) deploy/pull the container image to CloudRun
(4) update Website -> rebuild -> push again with new tag, and roll out new version of Website without Downtime (CI/CD)
(5) GKE
gcloud auth configure-docker
gcloud services enable []
gcloud builds submit --image=[] --plateform managed
gcloud run deploy --image=[] --plateform managed
ctrl+c
from Step 0
-
prework-1, clone a existing website code from Github
git clone https://github.com/googlecodelabs/monolith-to-microservices.git cd monolith-to-microservices cd ./setup.sh -
prework-2, test app by starting the web server
cd ~/monolith-to-microservices/monolith npm start [output] Monolith listening on port 8080! [output in UI]
-
prework-3, stop the Web server
pressing CTRL+C in Cloud Shell.
Dockerize App use Docker cmd line
-
1.1, in cloud shell, log in Docker
sudo usermod -a -G docker ${USER} //If you're using a Linux-based operating system, such as Ubuntu or Debian, add your username to the docker group so that you can run Docker without using sudo. // Caution: The docker group is equivalent to the root user. See Docker's documentation for details on how this affects the security of your system. // Log out and log back in for group membership changes to take effect. If you are using a virtual machine, you may need to restart the virtual machine for membership changes to take effect. -
1.2, test Docker run within output of Date
docker run busybox date -
1.3, log in the repo
gcloud auth configure-docker -
1.4, pull the docker image to repo
docker pull gcr.io/google-samples/hello-app:1.0 -
1.5, tag the docker image and push it to repo
docker tag gcr.io/google-samples/hello-app:1.0 gcr.io/PROJECT_ID/[tag name] docker tag gcr.io/google-samples/hello-app:1.0 gcr.io/PROJECT_ID/quickstart-image:tag1 docker push gcr.io/PROJECT_ID/quickstart-image:tag1 -
1.6, on contrast, we pull back the app from remote to local.
docker pull gcr.io/PROJECT_ID/quickstart-image:tag1 -
1.7, delete docker image in repo
gcloud container images delete gcr.io/PROJECT_ID/quickstart-image:tag1 --force-delete-tags
// in project id includes : colons symbol, then check out the doc https://cloud.google.com/container-registry/docs/overview#domain-scoped_projects
-
1.8, sum up doc
from step 3
Dockerize App use Cloud Build
-
2.1, in cloud shell, compress and move docker file to Cloud Sorage Bucket.
-
2.2, enable google cloud api called cloudbuild
gcloud services enable cloudbuild.googleapis.com -
2.3, use Cloud Build to create container, and push it to container registry.
gcloud builds submit --tag gcr.io/${GOOGLE_CLOUD_PROJECT}/monolith:1.0.0 . [output] ----------------------------------------------------------------------------------------------------- ------------------------------------------------------------------------------------------------------ ID CREATE_TIME DURATION SOURCE IMAGES STATUS 1ae295d9-63cb-482c-959b-bc52e9644d53 2019-08-29T01:56:35+00:00 33S gs://<PROJECT_ID>_cloudbuild/source/1567043793.94-abfd382011724422bf49af1558b894aa.tgz gcr.io/<PROJECT_ID>/monolith:1.0.0 SUCCESS -
2.4, in cloud concole, navigation bar > Cloud Build > history
from step 3
Deploy a Container
-
3.1, enalbe cloud run service
gcloud services enable run.googleapis.com -
3.2, deploy it to Cloud Run PaaS
gcloud run deploy --image=gcr.io/${GOOGLE_CLOUD_PROJECT}/monolith:1.0.0 --platform managed -
3.3, mgmt the Deployment
// 1. choose the managed version of Cloud Run by specifying --platform managed. // 2. specify which region you'd like to run in. Type the number for the region closest to you. // 3. accept the default suggested service name (it will be "monolith") by pressing Enter. -
3.4, check the deployment
gcloud run services list Type "1" to choose the first option: [1] Cloud Run (fully managed) [the pod is running now] SERVICE REGION URL LAST DEPLOYED BY LAST DEPLOYED AT ✔ monolith us-east1 <your url> <your email> [time] -
3.5, setup Endpoint for App in Cloud Run
gcloud run deploy --image=gcr.io/${GOOGLE_CLOUD_PROJECT}/monolith:1.0.0 --platform managed --concurrency 1 // change the concurrency val as 1 -
3.6, in cloud console, check the result by navigating to Cloud Run > [Service Name] > revsion tab.
-
tips & attentions:
By default, a Cloud Run application will have a concurrency value of 80, meaning that each container instance will serve up to 80 requests at a time. This is a big departure from the Functions-as-a-Service model, where one instance handles one request at a time.
from step 4
Modify code of website, and deploy it again
prework,
cd ~/monolith-to-microservices/react-app/src/pages/Home
mv index.js.new index.js
cat ~/monolith-to-microservices/react-app/src/pages/Home/index.js
cd ~/monolith-to-microservices/react-app
npm run build:monolith
// copy it to monolith dir
-
4.1 rebuild the Docker container and publish it to Container Registry. You can use the same command as before, except this time you will update the version label.
cd ~/monolith-to-microservices/monolith npm start // feel free to do this execution gcloud builds submit --tag gcr.io/${GOOGLE_CLOUD_PROJECT}/monolith:2.0.0 . // with new version name this time -
4.2, deploy the new version of service.
gcloud run deploy --image=gcr.io/${GOOGLE_CLOUD_PROJECT}/monolith:2.0.0 --platform managed
from step 5
since the docker image is saved in the private registry, GKE can pull from it to do GKE-style deployment.
see https://github.com/QueenieCplusplus/GKE
66(container), 10445, 14743(go)